INDEX

Explanations

variability, low saturation, drops

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

<bos>

2.11

('--

1.96

りに

1.93

ुका

1.93

특별시

1.92

り

1.91

 روست

1.90

-_

1.90

్వ

1.88

 Büro

1.86

POSITIVE LOGITS

हाल

2.24

一份

2.24

 μιας

2.21

𝐢

2.15

concatenate

2.13

𝐨

2.12

িমা

2.10

ﻟ

2.09

 decks

2.09

𝐚

2.09

Activations Density 0.375%